Policy 26 (2009) 545–550

Contents lists available at ScienceDirect

Land Use Policy

journal homepage: www.elsevier.com/locate/landusepol

Tree farms: Driving forces and regional patterns in the global expansion of

Thomas K. Rudel ∗

Department of Human Ecology, Rutgers University, 55 Dudley Road, New Brunswick, NJ 08901, article info abstract

Article history: People have planted in rural places with increasing frequency during the past two decades, but the Received 21 November 2007 circumstances in which they plant and the social forces inducing them to plant remain unclear. While Accepted 7 August 2008 that produce for industrial uses comprise an increasing number of the plantations, most of the growth has occurred in Asia where plantations that produce wood for local consumption remain Keywords: important. Explanations for these trends take economic, political, and human ecological forms. Growth Forest plantations in urban and global markets for forest products, coupled with rural to urban migration, may spur the Forest transition conversion of fields into farms. Government programs also stimulate . These programs Tree farms Rural to urban migration occur frequently in nations with high population densities. Quantitative, cross-national analyses suggest that these forces combine in regionally distinctive ways to promote the expansion of forest plantations. In Africa and Asia plantations have expanded most rapidly in nations with densely populated rural districts, rural to urban migration, and government policies that promote tree planting. In the Americas and Oceania plantations have expanded rapidly in countries with relatively stable rural populations, low densities, and extensive tracts of land in pasture. If, as anticipated, the growing concern with global warming spurs further expansion in forest plantations in an effort to sequester carbon, questions about their social and ecological effects should become more pressing. © 2008 Elsevier Ltd. All rights reserved.

Introduction plantations producing wood for industrial purposes have grown especially rapidly (Del Lungo et al., 2001). Landowners who decide to plant trees engage in an unusual As these people planted orderly rows of trees on more land dur- form of agriculture. Compared to row crops, planted trees grow ing the last two decades of the 20th century, controversies over the better on sloped land, require significantly less labor, and yield effects of tree farms on rural societies and ecologies grew louder. very little financial return for the first 7 or 8 years after plant- Advocates for tree farms (Victor and Ausubel, 2000) pointed to the ing (Bentley, 1989). For these reasons landowners only create tree very high rates of achieved on model plantations farms under a select set of conditions. Forest plantations currently in the tropics like the Aracruz holding in (Marchak, 1995) make up only about 5% of the world’s forests, but they have been and argued that these ‘fastwood’ farms can deliver important eco- increasing rapidly in extent. Between 1980 and 2000 the extent of nomic and environmental services. By meeting market demands forest plantations increased sevenfold in developing countries (Del for paper products from , forest plantations lower the Lungo et al., 2001). A wide variety of landholders have established price for pulpwood and, in so doing, reduce the economic incen- plantations. Smallholders in Africa created woodlots behind their tives to log natural forests (Kohlin and Parks, 2001). The rapid rates houses (Fortmann and Bruce, 1988; Tiffen et al., 1994). Corporations of growth achieved on tropical tree plantations also enable them to have planted extensive tree farms in Australia, North America, and sequester appreciable amounts of carbon, enough that, if planted South America (Marchak, 1995; Barlow and Cocklin, 2003). Villages on a large enough scale, tropical tree farms could offset a substan- in mainland Southeast Asia have planted trees in nearby uplands tial amount of the carbon emissions from the consumption of fossil (Muller and Zeller, 2002), and state agencies have planted trees on fuels (Gibbard et al., 2005; Righelato and Spracklen, 2007). degraded lands in South Asia (Guha et al., 1984). This diverse assem- Critics see forest plantations as vehicles through which 19th blage of forest owners may be changing. Since 1980 large century colonial officers (Peluso, 1992; Guha, 2000); and more recently corporate officials (Potter and Lee, 1998; Rohter, 2004) have deprived poor rural peoples of access to lands and livelihoods. ∗ Tel.: +1 732 932 9169x317; fax: +1 732 932 6667. The plantations also despoil the local ecology through the intro- E-mail address: [email protected]. duction of fast growing exotic species like Eucalyptus rostrata and

0264-8377/$ – see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.landusepol.2008.08.003 546 T.K. Rudel / Land Use Policy 26 (2009) 545–550

Pinus caribaea that, among other things, absorb so much water that forest industries covet large tracks of land because they intend to plantations of these species markedly reduce local stream flows take advantage of economies of scale in producing wood for large, (Cossalter and Pye-Smith, 2003). Since 2000 people have begun to oftentimes, global markets. Amerindians still use these lands, and voice an additional set of concerns about the ecological effects of conflict ensues. More generally, forest transition theory predicts planting transgenic trees in plantations (Petermann et al., 2006). that societies with streams of rural to urban migrants and growing These debates about the merits and demerits of forest planta- numbers of prosperous urban consumers will create more exten- tions, while valuable, leave unanswered some important questions. sive tree plantations. The literature focuses on high profile plantations in Brazil, Chile, While the –globalization hypothesis explains the and Indonesia, but the representativeness of these plantations expansion of forest plantations largely in terms of market mecha- and the associated political conflicts remains open to question. nisms involving individuals, another line of theorizing focuses on The spatial and social characteristics of forest plantation expan- government actions and the aggregated environmental effects of sion can best be appreciated through global scale analyses like forest plantations. Not surprisingly, the driving forces in this expla- the one offered here. Statistical analyses of these patterns provide nation are political. Government officials decide that the ecosystem an opportunity to assess explanations for plantation expansion, services provided by forests are important to the well being of the one that emphasizes the economic and demographic forces that society, and they create programs to expand the size of forests by accompany urbanization, another that focuses on political initia- planting trees. The most commonly cited environmental service tives to increase environmental services by expanding the size provided by forests involves flood control. Trees absorb water, so, of forests, and a third one that emphases high population den- at least in theory, when the headwaters region of a largely forested sities in rural areas. To answer these empirical and theoretical river basin experiences heavy rains, the trees absorb the water and questions, this paper presents a cross-sectional analysis of the reduce the likelihood of downstream flooding. When the rains fall extent of forest plantations in 2005 followed by an analysis of on cleared land, the water runs off of it, and the increased volume of changes in the extent of forest plantations between 1990 and runoff causes flooding downstream (Bradshaw et al., 2007). In the 2005. aftermath of damaging floods politicians want to do something, so they have put programs in place to increase . If the new Theory programs expand local supplies of wood and reduce a country’s timber imports, so much the better from the politicians’ point of Social scientists have offered several different explanations for view. the increasing extent of forest plantations. Economists theorize Government-led programs of tree planting have occurred in a that the recent growth in forest plantations of all types stems variety of contexts. The Chinese government accelerated already from increases in the scarcity of forest products throughout the massive efforts at tree planting in the aftermath of the great Yangtze world (Foster and Rosenzweig, 2003). At least one body of the- River floods in the late 1990s (Mather, 2007). The federal gov- ory, referred to as forest transition theory, draws upon geography ernment in the United States initiated a Shelterbelt Project of and sociology as well as economics to explain the recent expansion planted trees to forestall further dust storms in the Great Plains in forest plantations. Forest transition theory would attribute the during the 1930s (Droze, 1977). Non-governmental organizations expansion of forest plantations to shifts in market forces that occur (NGOs) with similar goals of environmental improvement fre- with the urbanization of societies and the globalization of forest quently complement and strengthen these state led efforts at products markets (Rudel et al., 2005). The concentration of people landscape transformation (Brechin, 1997). This sequence of events in cities raises their incomes and increases their expenditures on may take a new form in the 21st century if and when heads of state wood products, so markets for forest products grow. The accompa- decide that they want to take advantage of the carbon sequester- nying rural to urban migration makes farm labor more expensive ing environmental services provided by planted forests in order to and compels farmers to reduce the amount of land that they farm. combat global warming. Farmers respond to these labor and scarcities by A final line of theorizing about the changing incidence of forest converting their least productive fields, often on sloped lands, into plantations emphasizes variations in the human ecology of rural tree farms which require less labor than row crops (Mather and areas. In some places like central high population densi- Needle, 1998; Schirmer et al., 2005). In instances where the chil- ties in rural areas have created pressures to eliminate forests for dren have left a farm, the conversion from crops to trees may not cropped land (Williams, 2006, pp. 310–311). Alternatively, arid con- occur until the land passes into the hands of a younger generation ditions may have in some places caused cultivators to eliminate that does not want to farm. In this sense the decline in the fam- forests in the few places where they could grow naturally. Under ily farm presages the rise of the forest plantation (Neumann et al., these circumstances wood products grow scarce, and people may 2007). respond by planting trees. These processes may explain why the Where rural to urban migration has not ‘emptied out’ rural longest lived government assisted tree planting programs often areas, the creation of forest plantations often leads to conflict. occur in countries with arid regions, like or countries in Nineteenth century colonial regimes expropriated lands held by the Sahel (Brechin, 1997). This line of reasoning explains differ- smallholders in order to secure supplies of wood and induced ential rates of plantation expansion as the outcome of coupled widespread resistance (Peluso, 1992; Guha, 2000). More recently, human–natural situations that vary both regionally and histori- when a regional development agency induced smallholders in cally. Lesotho to fence off some grazing land for woodlots in the 1980s, Data collected by officials at the Food and Agricultural Orga- pastoralists tore down the fences and uprooted the trees (Ferguson, nization of the United Nation (FAO) on the changing extent of 1990, pp. 242–243). Some of the most protracted conflicts over forest plantations across nations makes it possible to assess, forest plantations in Latin America have involved corporations through statistical analyses, the merits of the market-based and Amerindians like the Mapuche of southern Chile (Rohter, urbanization–industrialization explanation, the more political 2004). Amerindians have enjoyed particularly rapid rates of nat- environmental services explanation, and the human ecologi- ural increase in Latin America over the past two decades and they cal explanation for recent increases in the extent of forest are less likely to migrate to urban areas than their mestizo neigh- plantations. The methods of analysis and results are outlined bors (McSweeney and Arps, 2005). Under these circumstances, below. T.K. Rudel / Land Use Policy 26 (2009) 545–550 547

Materials and methods The analysis includes nations from both the global north and south. A lack of data made it impossible to include two relatively The data for the following analysis comes from the Food and small categories of nations in the analyses. Island nations with few Agricultural Organization of the United Nations. Beginning in 1990, people and limited resources frequently do not report data on land FAO officials began requesting data on the extent of planted forests cover change to FAO, so they are not included in the analysis. Former from officials in each country. They repeated their requests for Soviet states that became independent nations after 1990 are not the Forest Resource Assessments of 2000 and 2005. To make the included in the analysis because demographic data on these places data comparable, FAO officials established a common definition for the 1980s could not be obtained. Eastern European nations that for use in estimating the extent of planted forests (FAO, 2004). underwent a change in regime during the same period are included For each country they conducted a detailed and in the analysis. examined changes in national forest policy. I use these data to The multivariate analyses of the data take two forms. Ordinary measure changes in the extent of forest plantations and changes least squares (OLS) regression analyses of landscape change weight in forest policy pertaining to the plantations. Two features skew political jurisdictions equally. In other words in these unweighted estimates from the FAO data in non-random ways. First, the FAO analyses large continental countries like the United States count definition of forest plantations includes larger planted areas (>.5 as much as small island nations like the Bahamas. Weighting the hectares), but misses many small woodlots behind smallhold- multivariate analyses by the size of a country’s land area gives ers’ homes. This omission leads, most likely, to underestimates a somewhat different picture about the prevalence of particular of the extent of planted forests in Africa because woodlots com- patterns of landscape change. Large countries like , China, prise a large proportion of the planted forests, particularly in India, Russia and the United States have an influence in these anal- Western and Southern Africa (Holmgren et al., 1994; Tiffen et al., yses commensurate with their size. I report both unweighted and 1994). Second, affects the extent of forest plan- weighted analyses in the following tables. Taken together, the two tations in some places. Countries with humid forest areas, like different multivariate techniques provide different but related per- Canada, tend to have fewer plantations because natural regener- spectives on processes of forest plantation expansion around the ation repopulates many cutover areas. in these countries globe. often manage ‘modified’ natural forests rather forest plantations. Weighted analyses have an additional methodological advan- While these measurement issues detract from the value of the tage. With the data for each nation derived from government data as the basis for an analysis, they need to be assessed against records, there will be variations in the amount of error from the considerable value of a data set that is both historical and nation to nation. Some nations will have accurate cadastral records global in its reach. This assessment suggests going ahead with indicating the extent of forest plantations while others will not. an analysis, acknowledging all the while that it is exploratory in Under these circumstances there might be problems of het- nature. eroscedasticity (unequal error variances) in the data (Berry and Table 1 presents the measures and the sources for the data used Feldman, 1985, p. 74). An inspection of the residuals plotted in the analyses that follow. If the forest transition hypothesis accu- against values of the independent variables does not indicate the rately describes the drivers of plantation expansion, then plantation kind of skewed distributions in the residuals that would indi- expansion should occur in wealthier societies with declining rural cate heteroscedasticity. Nevertheless, the weighted least squares populations. The political hypothesis would receive support if the analytic approach provides a useful check on the OLS anal- presence or absence of a government policy to encourage forest yses because, like generalized least squares (GLS) estimation plantations in a country makes a discernible difference in rates of techniques, weighted least squares analyses produce unbiased esti- forest plantation expansion. A human ecological explanation for mators in the presence of heteroscedasticity (Berry and Feldman, forest plantation expansion would receive support if population 1985). The independent variables have condition indices of below density variables and related measures of forest per capita, the 3 which indicate that the equations do not suffer from seri- extent of cultivated land, and the extent of land in pasture figure ous problems of multi-collinearity. As a result, the equations in centrally in explanations for the increasing extent of forest planta- Tables 3–5 all seem to be free of serious contaminating influ- tions. ences.

Table 1 Measures and sources for data

Extent of forest plantations, 2005 Extent of planted forests in a country in 2005 (in hectares). Source: Global Forest Resource Assessment, 2005 (FRA2005), FAO. Accessed at http://www.fao.org/forestry/site/fra2005/en/ Change in forest plantations, 1990–2005 Planted forest in 2005—Planted forest in 1990. Source: Global Forest Resource Assessment, 2005 (FRA2005); accessed at http://www.fao.org/forestry/site/fra2005/en/ Cultivated land, 1990 Cultivated land as a Percent of Land Area (crops and pasture), 1990. Source: Resource Stat, FAOSTAT; accessed at http://faostat.fao.org/site/377. Forest plantation policy Policy to Encourage the Creation of Forest Plantations. From Country Profiles at http://www.fao.org/forestry/site/fra2005/en/ Percent in pasture, 1990 Percent of Land Area in Pasture, 1990. Source. Resource Stat, FAOSTAT; accessed at http://faostat.fao.org/site/377. Population density, 1990 Population, 1990/Area of Country. Sources: Penn World Tables at http://pwt.econ.upenn.edu/php site/pwt62/pwt62 form.php. World Bank at http://web.worldbank.org/WBSITE/EXTERNAL/DATASTATISTICS. Forest per capita, 2000 Extent of Closed Forests, 2000/Population, 2000. Source: Global Forest Resource Assessment, 2005 (FRA2005), FAO. Accessed at http://www.fao.org/forestry/site/fra2005/en/. Percent forested, 1990 Percent of Land Area in Forests, 1990. Source: Global Forest Resource Assessment, 2005 (FRA2005), FAO. Accessed at http://www.fao.org/forestry/site/fra2005/en/. Percent change in rural population, 1980–1990 Rural Population, 1980—Rural Population, 1990. Divided by Rural Population, 1980. Sources: United Nations, Department of International Economic and Social Affairs, 1985 Estimates and Projections of Urban, Rural, and City Populations. United Nations, 1992. Human Development Report. GNP per capita (logged), 1990 Source: Penn World Tables at http://pwt.econ.upenn.edu/php site/pwt62/pwt62 form.php. 548 T.K. Rudel / Land Use Policy 26 (2009) 545–550

Table 2 Regional patterns of change in forest plantations

Region Forest plantations as a percentage of Extent of forest plantations, 2005 Change in extent of forest plantations, all forests in a country (thousands of hectares) 1990–2005 (thousands of hectares)

Africa—Sub-Saharan 7.6 314 17 Asia—South and East 12.5 7605 3130 Asia–Central 10.8 243 .2 Europe—Northern 12.6 847 196 Mediterranean Basin (Southern Europe and 26.7 509 135 North Africa) North America 4.9 1449 584 Oceania 10.5 768 346 South America 6.8 1032 272

Source: Food and Agriculture Organization of the United Nations. FRA2005 Data base.

Results countries with low population densities, few policy directives, and little cultivated land still had extensive forest plantations. Brazil, Table 2 describes regional variations in forest plantation expan- Russia, and the United States would all fit this profile. sion over a 15-year period between 1990 and 2005. The expansion Tables 4 and 5 present analyses of changes in the extent of for- concentrated in South and East Asia; by 2005 countries in this est plantations between 1990 and 2005. In the overall unweighted region contained 62% of the world’s recorded forest plantations. analysis in column 1 of Table 4 only the control variable, percent of People planted trees in a wide range of other settings. Only Africa area in tree farms in 1990, associates significantly with changes in and Central Asia saw little change in the extent of forest plantations the extent of tree farms. In the weighted analysis in column 2 pop- and, as noted above, the pattern in Africa may reflect problems with ulation density in 1990 and the population density–forest policy the data rather than an absence of tree planting in the region. The interaction term associate positively with plantation growth. For- importance of forest plantations as a source of wood also varies est per capita in 1990 associates negatively with plantation growth. regionally. While forest plantations only constitute 5% of all forests These findings most likely reflect the tremendous increase in the in the Americas, they represent more than 25% of all forests in the extent of forest plantations in densely populated places in South Mediterranean basin. The beneficiaries from expansion Asia (India) and East Asia (China and ) after 1990 (Mather, also vary regionally. In the Americas large landowners and corpo- 2007). rations owned 67% of the forest plantation land in 2000, so most Table 5 explores the regional patterns of plantation expansion. of the economic benefits from plantations went to already wealthy The findings for Africa in columns 1 and 2 suggest the importance interests. In Asia private landowners running large industrial plan- of densely settled landscapes with large proportions of cultivated tations only controlled 5% of all plantation land in 2000. States land, coupled with forest plantation promotion policies, in the and villages controlled most plantation land. In addition people expansion of forest plantations. In the findings for Asia–Europe in used large numbers of plantations for non-industrial purposes, as columns 5 and 6 rural to urban migration predicts forest plantation sources for fuelwood and as protection for water sources (Del Lungo expansion (unweighted analysis), as forest transition theory would et al., 2001). Given these different patterns of ownership and use, predict, but, like Africa, Asia–Europe also features a significant pop- expansion in the extent of plantations should have quite different ulation density–policy interaction term (weighted analysis). The distributive consequences in South and East Asia than it does in the association between plantation growth and a low proportion of Americas. cultivated land in the same equation probably reflects conditions The analyses reported in Tables 3–5 begin to identify the social in Russia and the larger European nations. In the findings for the forces behind plantation expansion. The cross-sectional multi- Americas and Oceania in columns 3 and 4 rural population stability, variate analyses in Table 3 indicate in both the weighted and contrary to forest transition theory’s predictions, predicts planta- unweighted analyses that densely populated nations with little tion growth (weighted analysis) as does extensive tracts of pasture forest cover and considerable income have relatively extensive land in 1990 (unweighted analysis). forest plantations in 2005. China exemplifies this pattern. The negative coefficients for cultivated land and for the population density–forest policy interaction term indicates that some large Table 4 Determinants of growth in the extent of forest plantations, 1990–2005 Table 3 Variations in the extent of forest plantations across nations, 2005 Unweighted Weighted analysis analysis (by size of country) Unweighted Weighted analysis % of land area in forest .736* (.381) analysis (by country size) plantation, 1990 Population density, 1990 .022*** (.004) .047*** (.006) % change in rural population, .002* (.001) Population density, 1990* −.019*** (.005) 1980–1990 Forest Plantation Policy Population density, 1990 * .043*** (.011) Forests per capita, 1990 −.092*** (.028) −.077** (.024) plantation forest policy Cultivated land as % of land area −.074*** (.022) Plantation forest policy .039 (.026) GNP per capita, 1990 .005* (.002) .002* (.001) Cultivated land as % of land area .113 (.088) Forests per capita, 1990 −.148* (.089) R2 (adj.) .270 .420 N of cases 114 114 R2 (adj.) .055 .128 N of Cases 112 112 Note: * = <.10, ** = <.01, *** = <.001. The numbers reported in the same row with the variables are unstandardized regression coefficients. Their standard errors are Note: * = <.10, ** = <.01, *** = <.001. The numbers reported in the same row with immediately underneath the coefficients. Regressions run on cross-sectional data the variables are unstandardized regression coefficients. Their standard errors are for 1990 produces quite similar results. immediately underneath the coefficients. T.K. Rudel / Land Use Policy 26 (2009) 545–550 549

Table 5 Regional patterns in forest plantation expansion, 1990–2005

Africa Americas and Oceania Europe and Asia

Unweighted Weighted Unweighted Weighted Unweighted Weighted

% Land in forest plantations, 1990 .019** (.006) .031*** (.008) 4.947*** (.739) Population density 1990 .041* (.015) Population density 1990 * forest policy .092* (.037) .057* (022) Rural population change, 80–90 .003** (.001) −.016* (.008) −.005 (.004) Cultivated land as % of land 1990 .005*** (.001) −.001 (.001) −.257* (.122) Forests per capita, 1990 −.190 (.142) % pasture, 1990 .003** (.001)

R2 (adj.) .511 .267 .480 .685 .074 .129 N of Cases 38 38 27 27 47 47

Note: * = <.10, ** = <.01, *** = <.001. The numbers reported in the same row with the variables are unstandardized regression coefficients. Their standard errors are immediately underneath the coefficients.

Discussion abandoned uplands with commercially valuable species of trees (Mather, 2007). In all three countries the presence of millions One of the more interesting findings in Tables 3 and 4 involves of smallholders on the lands magnified the impact of the policy the changed sign in the coefficient for the population den- reforms, creating the observed interaction effect, on plantation sity × forest policy interaction term. In the analysis of the extent of expansion. This regional trajectory in land use change appears to forest plantations across nations in 2005, the coefficient is negative, characterize South and East Asia as well as parts of Southern and indicating that many European and American polities with either Eastern Africa. low population densities or little policy direction regarding tree The pattern of low population densities, no policy initiatives, and farms, nonetheless, have extensive tree farms. In Table 4 the sign still substantial gains in forest plantations in the cross-sectional switches. Recent increases in forest plantations occurred primarily analyses would appear to describe rural regions in the Americas in densely populated places with effective policies for promoting with extensive land uses (pastures), relatively stable rural popula- the expansion of forest plantations. China, India, and Vietnam, all tions, and large urban markets for wood. Because forest plantations of which promoted forest plantation expansion through a variety in these places usually produce wood for industrial uses in local of political reforms enacted during the 1980s and 1990s, exem- and global markets (Marchak, 1995), their managers want to pro- plify this pattern of forest plantation expansion (Mather, 2007). duce wood on extensive tracts of land. This expansionist agenda These regional contrasts point to the changing historical geogra- sometimes places these managers on collision courses with the phy of plantation growth. While forest plantations first became leaders of stable or expanding populations of rural residents, laying fairly extensive in European and American countries that currently the bases for the political controversies surrounding forest planta- have relatively stable rural populations, most of the recent growth tions in Chile and Brazil. These large scale forest plantations and in plantations has occurred in the densely populated, extensively the associated pattern of change would appear to predominate in cultivated South and East Asian countries that have experienced the Americas, Russia, and Oceania. substantial rural to urban migration during the past 30 years. There are two regional trajectories and no single explanation The results from the Asia–Europe analyses both confirm and for the current patterns of forest expansion. One trajectory, in challenge forest transition theory. As forest transition theory Asia, Europe, and Africa, features dense populations of smallholders would suggest, rural to urban migration in this context appears who plant trees, frequently with government assistance. The other, (unweighted analysis, column 5, Table 5) to expedite the creation of in the Americas, Russia, and Oceania, involves larger plantations, tree farms because the emigrants abandon lands that then become sometimes under corporate ownership. These patterns are prob- the sites for forest plantations. Neither forest transition theory ably best understood through a human ecological approach that nor the political explanation anticipates the importance of high incorporates and uses insights from both forest transition theory population densities, in combination with political initiatives, in and state led theories of political change. promoting the expansion of tree farms in both Africa and Asia. Events in China, India, and Vietnam illustrate the dynamics of this Conclusion interaction (Mather, 2007). In China a series of state led afforesta- tion campaigns during the 1980s extended plantation forests, but Questions about political dimensions appear particularly press- the rate of expansion accelerated after the government introduced ing in setting agendas for future research on forest plantation the ‘grain for green’ program in the aftermath of the Yangtze River expansion. Given the increasing political salience of global warming floods in 1997 that gave small farmers financial incentives if they and the ability of forests to sequester carbon, more state led efforts converted some of their fields into forest plantations. In India to increase the extent of forest plantations seem likely. While the central government-led programs failed to end and contribution of carbon sequestering forest plantations to the over- degradation during the 1980s. The government then decentralized all solution of the global warming problem may be small relative to control over many forests, giving villagers a share of the proceeds reductions in emissions of fossil fuels and declines in the continu- from the sale of local forest products. Under these circumstances ing destruction of natural forests (Cossalter and Pye-Smith, 2003; villagers begin to rehabilitate and restore forests (Poffenberger and CIFOR, 2007), forest plantation expansion is one of the few pol- McGean, 1996). In Vietnam devolution of some control over the icy levers to counter global warming that would reinforce a trend proceeds from the land, coupled with an improved transportation already under way and therefore would be relatively easy to imple- network, led some villages and smallholders to abandon unproduc- ment. Further research might address questions about the effects tive upland agriculture and rely on purchased foodstuffs from more of different kinds of state interventions on forest cover, environ- agriculturally productive areas of the country. Villagers planted the mental services, and rural society. Mather (2007) has provided a 550 T.K. Rudel / Land Use Policy 26 (2009) 545–550 model for this type of inquiry in his recent comparative study of Ferguson, J., 1990. The Anti-Politics Machine: Development, Depoliticization, and forest reforms in China, Vietnam, and India. Bureaucratic Power in Lesotho. Cambridge University Press, Cambridge, England. Food and Agricultural Organization of the United Nations (FAO), 2004, Global Forest This research agenda presents both a challenge and an oppor- Resources Assessment Update 2005: Terms and Definitions. Rome. Accessed at tunity for social scientists and policymakers. Several of the regions http://www.fao.org/forestry/site/fra2005-terms/en/. with the most climactic potential for (Sub- Fortmann, L., Bruce, J., 1988. Whose Trees? Proprietary Dimensions of Forestry. Westview Press, Boulder, CO. Saharan Africa, South, Southeast, and East Asia) also contain the Foster, A., Rosenzweig, M., 2003. Economic growth and the rise of forests. The Quar- worlds’ largest concentrations of poor people. Under these circum- terly Journal of Economics 118, 601–637. stances payments for carbon sequestration from forest plantations Gibbard, S., Caldeira, K., Bala, G., Philips, T., Wickett, M., 2005. Climate effects could provide a valuable supplemental source of income for the of global land cover change. Geophysical Research Letters 32 (23), Art. No. L23705. rural poor. Efforts to realize this type of sustainable development Guha, R., 2000. The Unquiet : Ecological Change and Peasant Resistance in must contend with two recurring patterns. First, the carbon seques- the Himalaya. University of California Press, Berkeley, CA. tering plantations must not be carved out of existing rain forests, Guha, R., Prasad, S., Gadgil, M., 1984. Deforestation and degradation of natural plant resources in India. Journal of Indian Anthropology and Sociology 19, as has frequently occurred in the outer islands of Indonesia (Pirard 246–253. and Cossalter, 2006). Otherwise, the emissions from deforestation Holmgren, P., Masakha, E., Sjoholm, H., 1994. Not all African land is being degraded: nullify the gains from later sequestration. Second, states and corpo- a recent survey of trees on farms in Kenya reveals rapidly increasing forest resources. AMBIO 23, 390–395. rations have managed most large forest plantations, in part because Kohlin, G., Parks, P., 2001. Spatial variability and disincentives to harvest: deforesta- only they have the capital and secure land tenure necessary to cre- tion and fuelwood collection in South Asia. Land Economics 77, 206–218. ate the plantations and in part because they have been reluctant Lee, R., Field, D. (Eds.), 2005. Communities and Forests: Where People Meet the Land. Oregon State University Press, Corvallis, OR. to share profits with the poor who live around the plantations. For Marchak, P., 1995. the Globe. McGill-Queens University Press, Montreal. the rural poor of these regions to benefit from carbon sequestering McSweeney, K., Arps, S., 2005. ‘A demographic turnaround’: the rapid growth of plantations, they must secure tenure to these lands. Zimmerman indigenous populations in lowland Latin America. Latin American Research Review 40 (1), 3–29. (1951,p.376,quotedinLee and Field, 2005, p. 24) underlined the Mather, A., 2007.Recent Asian forest transitions in relation to forest transition theory. importance of this condition more than 50 years ago. International Forestry Review 9, 491–502. Mather, A., Needle, C., 1998. The forest transition: a theoretical basis. Area 30, “All perennial culture, but particularly the planting of trees, rests 117–124. on the stability of social institutions. No one would be foolish Muller, D., Zeller, M., 2002. Land use dynamics in the central highlands of Vietnam: enough to spend a decade or more ...to build up an olive grove, a spatial model combining village survey data with satellite imagery interpre- tation. Agricultural Economics 27, 333–354. which can bear fruit for a century unless he feels reasonably Neumann, P., Krahn, H., Krogman, N., Thomas, B., 2007. My grandfather would role sure of a reward for himself and his descendants.” over in his grave’: family farming and tree plantations on farmland. Rural Soci- ology 72, 111–135. States and NGOs can help the rural poor reap benefits from car- Petermann, A., Carman, N., Smith, D., 2006. Social and environmental impacts: forest bon sequestering forest plantations by helping them secure tenure plantations and genetically engineered trees. In: Paper presented at the 2006 IUFRO (International Union of Forest Research Organizations) meeting on Forest in land, negotiate payment for environmental services arrange- Plantations, Charleston, SC. ments, and obtain the capital to plant the carbon absorbing trees Peluso, N., 1992. Rich Forests, Poor People: Resource Control and Resistance in Java. (Boyd et al., 2007). University of California Press, Berkeley, CA. Poffenberger, M., McGean, B. (Eds.), 1996. Village Voices, Forest Choices: Joint in India. Oxford University Press, Delhi. References Potter, L., Lee, J. 1998. Tree Planting in Indonesia: Trends, Impacts, and Directions. Occasional Paper #18. Center for International Forestry Research (CIFOR), Bogor, Barlow, K., Cocklin, C., 2003. Reconstructing rurality and community: plantation Indonesia. forestry in Victoria, Australia. Journal of Rural Studies 19, 503–519. Pirard, R., Cossalter, C. 2006. The Revival of Industrial Forest Plantations in Indone- Bentley, J., 1989. Bread forests and new fields: the ecology of and forest sia’s Kalimantan Province: Will they help eliminate fiber shortfalls at Sumatran clearing among small woodland owners in . Journal of Forest History 17, Pulp Mills or Feed the China Market? Working paper #37, Center for Interna- 188–195. tional Forestry Research (CIFOR), Bogor, Indonesia. Berry, W., Feldman, S., 1985. Multiple Regression in Practice. Sage Publications, Bev- Righelato, R., Spracklen, D., 2007. Carbon mitigation by biofuels or by saving and erly Hills, CA. restoring forests. Science 317, 902. Boyd, E., Gutierrez, M., Chang, M., 2007. Small scale forest carbon projects: adapting Rohter, L. 2004. Mapuche Indians struggle to take back forests. The New York Times. CDM to low income communities. Global Environmental Change 17, 250–259. August 11, 2004. Bradshaw, C., Sodhi, N., Peh, K., Brook, B., 2007. Global evidence that deforestation Rudel, T., Coomes, O., Moran, E., Achard, F., Angelsen, A., Xu, J., Lambin, E., 2005. amplifies flood risk and severity in the developing world. Global Change Biology Forest transitions: towards a global understanding of land use change. Global 13 (11), 2379–2395. Environmental Change 15, 23–31. Brechin, S., 1997. Planting Trees in the Developing World. Johns Hopkins University Schirmer, J., Parsons, M. Charalambou, C., Gavran, M. 2005. Socio-economic impacts Press, Baltimore. of plantation forestry in the Great Southern region of western Australia, Center for International Forestry Research (CIFOR), 2007, CarboFor: Reducing Emis- 1991–2004. Report #PN04.4007. Forest and Wood Products Research and Devel- sions from Deforestation. Available at: http://www.cifor.cgiar.org/carbofor/. opment Corporation, Melbourne. Cossalter, C., Pye-Smith, C., 2003. Fastwood Forestry: Myths and Realities. Center for Tiffen, M., Mortimore, M., Gichuki, F., 1994. More People, Less Erosion: Environmen- International Forestry Research, Bogor, Indonesia. tal Recovery in Kenya. John Wiley and Sons, Chichester. Del Lungo, A., Vuorinen, P., Carle, J., 2001. Preliminary Analysis of Global Trends in Victor, D., Ausubel, J., 2000. Restoring the forests. Foreign Affairs 79 (6), 127–144. Forest Plantation Development, 1980–2000. Forestry Division, FAO, Rome. Williams, M., 2006. Deforesting the Earth: From Prehistory to Global Crisis. Univer- Droze, W., 1977. Trees, Prairie, and People: A History of Tree Planting in the Plains sity of Chicago Press, Chicago. States. Texas Woman’s University Press, Denton, TX. Zimmerman, E., 1951. World Resources and Industries, rev. ed. Harper, New York.